All of lore.kernel.org
 help / color / mirror / Atom feed
From: Colin D Bennett <colin@gibibit.com>
To: The development of GNU GRUB <grub-devel@gnu.org>
Cc: phcoder@gmail.com, Robert Millan <rmh@aybabtu.com>
Subject: Re: [URGENT] Re: bazaar
Date: Tue, 24 Nov 2009 07:50:29 -0800	[thread overview]
Message-ID: <20091124075029.34723a43@svelte> (raw)
In-Reply-To: <4B09E31C.7070207@gmail.com>

[-- Attachment #1: Type: text/plain, Size: 2174 bytes --]

On Mon, 23 Nov 2009 02:19:24 +0100
Vladimir 'φ-coder/phcoder' Serbinenko <phcoder@gmail.com> wrote:

> Carles Pina i Estany wrote:
> > Hello,
> >
> > I have to go to sleep now, but I've realized about a mistake or at
> > least unexpected behaviour.
> ...
> Looks like the mess is actually more profound than you describe.
> You
> replaced our mainstream with your branch. Looks like bzr failed at its
> primary task: protect against unintentional or intentional  deletion
> of files.  We should think of a way to make trunks and experimental
> branch commit-only. Meanwhile nobody pushes until further notice.
> I'll see how it can be recovered and protected.
> Robert: What is your latest backup before this accidental replacement?

The ‘append_revisions_only’ option should probably be set for all
public branches to prevent accidental non-commit changes (i.e.,
pushing that would reorder revisions, since ‘push’ is really a
mirroring operation and can change revision numbers for existing
revisions).  Add ‘append_revisions_only = True’ to the
branch's .bzr/branch/branch.conf (or, for new branches, use the
--append-revisions-only option when creating the branch with bzr init).
See
<http://doc.bazaar-vcs.org/bzr.2.0/en/user-reference/#append-revisions-only>
 for details.

When append_revisions_only is set, pushing to trunk can be done, but
it won't alter existing revisions, so it will prevent problems.  You
can't use the 'uncommit' command then, of course, but that is by
design.  To revert a revision (e.g., revno 44) you would use 
'bzr merge -r 44..43 ; bzr commit' instead, as this is an appending
operation only.

Obviously a malicious user can corrupt the repository since the
SSH/SFTP transport we're using allows full raw file access for
committers, but it's the accidental mess-ups we are really concerning
ourselves with since we have to trust committers anyway--hopefully we
have frequent automated backups.  (It might be worthwhile looking into
using revision signing at some point though as well as an extra measure
of security; I haven't used Bazaar's revision signing yet.)

Regards,
Colin

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 197 bytes --]

  parent reply	other threads:[~2009-11-24 15:50 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-11-23  0:44 bazaar Carles Pina i Estany
2009-11-23  1:19 ` [URGENT] bazaar Vladimir 'φ-coder/phcoder' Serbinenko
2009-11-23  1:40   ` [ISSUE SOLVED] " Vladimir 'φ-coder/phcoder' Serbinenko
2009-11-23  9:33     ` Robert Millan
2009-11-23  9:59       ` Carles Pina i Estany
2009-11-23 10:30         ` Robert Millan
2009-11-23 12:01       ` Vladimir 'φ-coder/phcoder' Serbinenko
2009-11-23  9:35     ` Robert Millan
2009-11-24 15:50   ` Colin D Bennett [this message]
2010-01-07 18:53     ` [URGENT] " Robert Millan

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20091124075029.34723a43@svelte \
    --to=colin@gibibit.com \
    --cc=grub-devel@gnu.org \
    --cc=phcoder@gmail.com \
    --cc=rmh@aybabtu.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.